LPS76 27April final.pub - Local Population Studies

advertisement
SOURCES AND METHODS
This item considers a range of sources and methods commonly used in local population
history. These vary in sophistication and complexity, but are intended to be of benefit to
the broad LPS readership, and are accompanied by worked examples. Each item is
written by an experienced population history practitioner, and will usually address both
the possibilities and the pitfalls of the respective sources and methods under discussion.
The LPS Board are happy to enter into correspondence on this item, which should be
addressed in the first instance to the LPS General Office.
THE COMPONENTS OF POPULATION CHANGE
Andrew Hinde
Andrew Hinde is Senior Lecturer in Population Studies at the University of
Southampton. He is the author of two books: Demographic methods (London, 1998) and
England’s population: a history since the Domesday Survey (London, 2003).
Demography is the study of population structure and change, and the most
fundamental quantity in demography is the total number of people living in a
particular place. This number will typically change over time, and how and
why it changes are often research questions of interest for local population
studies. Consider an example. In eastern Hampshire lie the small market town
of Alton and the neighbouring village of Holybourne.1 In the first census of
1801, the population of Alton was 2,026 persons and that of Holybourne was
366 (Table 1). During the next century the population of Alton increased
steadily so that by 1901 it was 5,479, more than two and half times what it was
in 1801. The population of Holybourne also increased, but by the third quarter
of the century the increase had stalled, and after 1871 its population began to
fall. This pattern of a rise and subsequent fall in population was characteristic
of many English villages during the nineteenth century.
Why did these two settlements have a different demographic history between
1801 and 1901? The ultimate answer lies in their different social and economic
evolution, as shaped by local, regional, national and even international forces.
Yet these forces can only act to change the population of a given place through
three processes: fertility, mortality and migration. It is through these
components of population change, and only through them, that the total number
of people living in a defined locality can change. In seeking to understand the
causes of population change, therefore, it is useful to try to measure how these
90
Table 1
Population of the Hampshire parishes of Alton and Holybourne, 1801–1901
Census year
Alton
Holybourne
1801
2,026
366
1811
2,316
384
1821
2,499
482
1831
2,742
487
1841
3,139
522
1851
3,538
497
1861
3,769
586
1871
4,092
618
1881
4,497
591
1891
4,671
587
1901
5,479
558
Source: G.S. Minchin, ‘Table of population’, in W. Page (ed.), Victoria history of Hampshire and the
Isle of Wight, 5 vols (London, 1912), vol. 5, 437.
three components have combined to produce the observed growth or decline
of the population.
The relationship between fertility, mortality and migration and population
change is encapsulated in what is often termed the demographic accounting
equation. If the population of a given place in a given census is denoted by the
symbol P1, and the population of the same place in the next census is denoted
by the symbol P2, then we can write down the following equation:
P1 + births – deaths + in-migrants – out-migrants = P2
where the ‘births’ and ‘deaths’, refer to the numbers of these events occurring
in the period between the two censuses, and ‘in-migrants’ and ‘out-migrants’
are the numbers of persons arriving and leaving the place over the same
period. The difference between births and deaths is called the natural increase of
the population, and the difference between the number of in-migrants and the
number of out-migrants is termed net migration. Since in the United Kingdom
there is not, and never has been, compulsory registration of changes of place of
residence, it is not normally possible to work out the numbers of in-migrants
and out-migrants.2 Therefore the demographic accounting equation is usually
simplified to read
P1 + births – deaths + net migration = P2.
The equation makes one important assumption: that the boundaries of the
place to which it is applied do not change between the two censuses. In other
words, it is important that the geographical extent of the area to which it is
91
applied is identical throughout the period being studied.
Consider the application of the equation to population change in the parishes
of Alton and Holybourne between 1871 and 1881. The population of Alton in
1871 was 4,092; by 1881 it had risen to 4,497 (Table 1). So we can write
4,092 + births – deaths + in-migrants – out-migrants = 4,497.
Similarly, for Holybourne (Table 1) we can write
618 + births – deaths + in-migrants – out-migrants = 591.
The next stage is to obtain the numbers of births and deaths. For certain
geographical units, such as registration districts (RDs) or counties, the numbers
of births and deaths in recent decades are published by the Registrar General.3
There were roughly 600 RDs in Victorian England; their average population
rose with the national population, but in 1881 it was about 20,000. For
individual parishes or small groups of parishes, though, civil registration data
on decadal numbers of births and deaths are not normally available.4 If
ecclesiastical registration is of good quality, the numbers of births and deaths
might be estimated from the numbers of baptisms and burials in the parish
registers. However, an alternative approach is to make the assumption that the
birth rate in the parish or parishes being studied was the same as that
prevailing in the RD in which the parish was situated. The annual number of
births and deaths in each parish can then be worked out by applying the RD
birth and death rates to the average population of the parish over the decade.
Both Alton and Holybourne were in the Alton RD, which during the 1870s had
a birth rate of 32.3 per thousand and a death rate of 16.7 per thousand, and an
average population of 15,098.5 The average population of Alton parish during
the decade 1871–1881 was (4,092 + 4,497)/2 which is 4,294.5. At a birth rate of
32.3 per thousand, 4,294.5 people will produce 4,294.5 x 0.0323 = 138.7 births
per year. Over the ten years between the 1861 and 1871 censuses, therefore, we
estimate that there were 1,387 births in Alton. Similarly, at a death rate of 16.7
per thousand, 4,294.5 people will generate an average of 71.7 deaths per year,
so that over the decade we can expect a total of 717 deaths. The natural
increase of the parish of Alton between 1871 and 1871 was therefore equal to
the difference between the decadal numbers of births and deaths, which is
1,387-717 = 670.
A similar calculation for the parish of Holybourne during the same decade
(average population 604.5), using the same birth and death rates, produces
estimates of 195 births and 101 deaths, resulting in a natural increase over the
decade of 94. Inserting these numbers into the demographic accounting
equations for the two parishes produces:
for Alton: 4,092+1,387-717+net migration = 4,497,
and for Holybourne: 618+195-101+net migration = 591.
The estimated net migration during the decade is then simply the number
92
Table 2
Components of population change in Alton and Holybourne, 1801–1901
Decade
Alton
Population
change
Natural
increase
1801–1811
290
1811–1821
183
1821–1831
Holybourne
Net
migration
Population
change
Natural
increase
Net
migration
274
16
18
47
-29
349
-166
98
63
35
243
377
-134
5
70
-65
1831–1841
397
362
35
35
62
-27
1841–1851
399
387
12
-25
59
-84
1851–1861
231
460
-288
89
77
12
1861–1871
323
499
-192
32
79
-47
1871–1881
405
670
-265
-27
94
-121
1881–1891
174
619
-445
-4
80
-84
1891–1901
808
406
301
-29
57
-86
Notes:
The natural increase figures in this table for decades before 1851 involve estimates of
either or both of the crude birth and death rates (see endnote 7).
Source:
G.S. Minchin, ‘Table of population’, in W. Page (ed.), Victoria history of Hampshire and
the Isle of Wight, 5 vols (London, 1912), vol. 5, 437.
required to make these equations balance. For Alton this is -265, and for
Holybourne it is -121. We estimate, therefore, that during the decade 1871–
1881, 265 more people left the parish of Alton to live elsewhere than arrived to
reside in Alton: for Holybourne, there were 121 more out-migrants than inmigrants.6
The growth in the population of Alton during the decade 1871–1881, therefore,
was produced by the natural increase of 670, but the effect of this was reduced
by about 40 per cent by net out-migration. In Holybourne, despite the natural
increase of 94, the population actually declined during the decade, because the
natural increase was more than cancelled out by heavy net out-migration.
Similar calculations can be applied to the remaining decades of the nineteenth
century for both parishes.7 The results (Table 2) show that for most decades of
the century out-migration exceeded in-migration in both parishes. Despite its
urban character as a market town, Alton only gained population through
migration in the decades 1801–1811, 1831–1841, 1841–1851 and 1891–1901, and
only in the last decade of the century was the increase through net migration
substantial. The village of Holybourne lost population quite heavily through
net migration in all decades except 1811–1821 and 1851–1861, with the rate loss
increasing during the last three decades of the century—a common pattern in
the villages of southern England.
It is important to recognise that, because we have used the same RD birth rates
and death rates to estimate the number of births in the two parishes, we have
assumed that fertility and mortality were the same in both. While this is clearly
93
unlikely to have been exactly true, there is plenty of evidence that variation in
fertility from place to place in nineteenth-century England was in general
rather small, and that major variations over short distances were uncommon.8
Local variations in mortality are more problematic, and it is probably not
advisable to apply RD-based death rates to parishes which are known to be
unrepresentative (for example, in their occupational structure) of the RDs in
which they are situated. In small units such as parishes, there will, of course,
also be random fluctuations in the numbers of births and deaths from year to
year, but over a whole decade these will tend to cancel out, so that decadal
totals of birth and deaths will exhibit much less random variation.
The demographic accounting equation can be applied at all geographical scales
from the individual parish, to groups of parishes, through registration districts
and counties to entire countries, provided that the boundaries of the units
analyzed do not change during the period of investigation. Generally speaking,
the reliability of the results increases with the size of the units, because
accurate data on births and deaths are more likely to be available.
Nevertheless, even at the parish level where the numbers of births and death
must often be estimated using birth rates which apply to larger geographical
units, the application of this method can provide useful estimates of the
contribution of fertility, mortality and migration to the observed growth or
decline of the population. This knowledge can then be related to what is
known of changes in the social and economic situation to aid understanding of
the causes of population change.
NOTES
1.
2.
3.
Alton and Holybourne are chosen here simply because the author resides close to the boundary
separating their respective parishes. Both, however, have their place in the cultural history of
England. It was in Alton on 24 August 1867 that an eight-year old girl called Fanny Adams was
murdered and her body horribly mutilated; her grave is in the town’s cemetery. Soon afterwards,
the expression ‘sweet Fanny Adams’ was adopted by sailors to describe some of the food they
were served at sea. Holybourne was the place where the novelist Elizabeth Gaskell, whose best
known works include Mary Barton: a tale of Manchester life (1848), North and south (1855) and Wives
and daughters (1866), died unexpectedly in 1865, in a house she had recently bought in the village.
Some progress in this direction can be made with nineteenth-century local data: see A. Hinde,
‘The use of nineteenth-century census data to investigate local migration’, Local Population Studies,
73 (2004), 8–28.
For the second half of the nineteenth century, for example, they are published in decennial
Supplements to certain of the Annual reports of the Registrar General of births, deaths and
marriages in England. These are in the British Parliamentary Papers (BPP) as follows: 1850s, BPP
1865/XIII [Cd. 3524]; 1860s, BPP 1875/XVIII, pt 2 [Cd. 1155–I]; 1870s, BPP 1884–5/XVII [Cd.
4564]; 1880s, BPP 1895/XXIII, pt 1 [Cd. 7769]; 1890s, BPP 1905/XVIII [Cd. 2619]. Death rates are
available by registration district (RD) for decades from 1841 onwards, and births for decades
from 1871 onwards. An alternative, more convenient, way to obtain RD births and deaths data
together with base populations for the decades from 1851 onwards is to use H. Southall and I.
Gregory, Great Britain historical database: census data: inter-censal population change statistics, 1851–
1961 [computer file], (London, 2004) which is available as study number 4556 from AHDS
History at the University of Essex (see www.ahds.ac.uk/history). Users based in UK HE/FE
institutions can download the tables from this study through the Collection of Historical and
Contemporary Census Data and Related Materials (see http://ahds.ac.uk/history/collections/
94
4.
5.
6.
7.
8.
chcc.htm). More details of electronic resources for the local population historians can be found in
H. Southall, ‘A vision of Britain through time: making sense of 200 years of census reports’, in
this issue of LPS, above, pp. 76–89.
It is possible to obtain totals of births and deaths for registration sub-districts (there were usually
two, three or four sub-districts within any one registration district) from the Annual reports of the
Registrar General. However, these are not presented on a decadal basis in the Supplements to the
Annual reports, neither are they available in a convenient electronic format (so far as the author is
aware), so using sub-districts involves a considerable amount of searching through the British
Parliamentary Papers. They do, however, permit more localised analysis.
British Parliamentary Papers, 1884–5/XVII, Registrar General of births, deaths and marriages in
England: supplement to forty-fifth annual report, 103.
It is important to recognise that knowing just the net migration, it is not possible to know exactly
how many in-migrants and out-migrants there were. For example, in the case of Alton, there
could have been 265 out-migrants and no in-migrants, or 965 out-migrants and 700 in-migrants,
or any other combination that produces a net migration of -265.
So far as the method described in this paper is concerned, it matters not how the registration
district (RD) birth and death rates are calculated, but simply that some birth and death rates
representative of the parishes being studied can be estimated. For those readers who are
interested however, the birth and death rates for the Alton RD in those decades when RD-specific
figures are not available were estimated by deriving a comparability factor based on the
relationship between the birth (or death) rate in the Alton RD and the national birth (or death)
rate during the decades for which RD-specific birth (or death) rates were available. This
comparability factor was used to adjust national rates for the decades before 1851 (in the case of
birth rates) and before 1841 (in the case of death rates) to provide estimated rates for the Alton
RD for those decades. The national birth and death rates were obtained from E.A. Wrigley and R.
S. Schofield, The population history of England 1541–1871: a reconstruction (Cambridge, 1989).
See, for example, R. Woods, The demography of Victorian England and Wales (Cambridge, 2000),
maps between pages 96–7.
95
Download